From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lf0-f69.google.com (mail-lf0-f69.google.com [209.85.215.69]) by kanga.kvack.org (Postfix) with ESMTP id 9327D280273 for ; Mon, 26 Sep 2016 13:28:31 -0400 (EDT) Received: by mail-lf0-f69.google.com with SMTP id y6so99495530lff.0 for ; Mon, 26 Sep 2016 10:28:31 -0700 (PDT) Received: from mx0a-001b2d01.pphosted.com (mx0b-001b2d01.pphosted.com. [148.163.158.5]) by mx.google.com with ESMTPS id xw9si20336128wjc.295.2016.09.26.10.28.29 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Mon, 26 Sep 2016 10:28:30 -0700 (PDT) Received: from pps.filterd (m0098420.ppops.net [127.0.0.1]) by mx0b-001b2d01.pphosted.com (8.16.0.17/8.16.0.17) with SMTP id u8QHMRnh098224 for ; Mon, 26 Sep 2016 13:28:29 -0400 Received: from e06smtp13.uk.ibm.com (e06smtp13.uk.ibm.com [195.75.94.109]) by mx0b-001b2d01.pphosted.com with ESMTP id 25q5duyq3m-1 (version=TLSv1.2 cipher=AES256-SHA bits=256 verify=NOT) for ; Mon, 26 Sep 2016 13:28:28 -0400 Received: from localhost by e06smtp13.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Mon, 26 Sep 2016 18:28:27 +0100 Received: from b06cxnps4074.portsmouth.uk.ibm.com (d06relay11.portsmouth.uk.ibm.com [9.149.109.196]) by d06dlp01.portsmouth.uk.ibm.com (Postfix) with ESMTP id 103A417D8056 for ; Mon, 26 Sep 2016 18:30:25 +0100 (BST) Received: from d06av03.portsmouth.uk.ibm.com (d06av03.portsmouth.uk.ibm.com [9.149.37.213]) by b06cxnps4074.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id u8QHSNP817695000 for ; Mon, 26 Sep 2016 17:28:23 GMT Received: from d06av03.portsmouth.uk.ibm.com (localhost [127.0.0.1]) by d06av03.portsmouth.uk.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id u8QHSMKt008417 for ; Mon, 26 Sep 2016 11:28:23 -0600 From: Gerald Schaefer Subject: [PATCH v4 2/3] mm/hugetlb: check for reserved hugepages during memory offline Date: Mon, 26 Sep 2016 19:28:10 +0200 In-Reply-To: <20160926172811.94033-1-gerald.schaefer@de.ibm.com> References: <20160926172811.94033-1-gerald.schaefer@de.ibm.com> Message-Id: <20160926172811.94033-3-gerald.schaefer@de.ibm.com> Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: Naoya Horiguchi , linux-mm@kvack.org, linux-kernel@vger.kernel.org, Michal Hocko , "Kirill A . Shutemov" , Vlastimil Babka , Mike Kravetz , "Aneesh Kumar K . V" , Martin Schwidefsky , Heiko Carstens , Rui Teng , Dave Hansen In dissolve_free_huge_pages(), free hugepages will be dissolved without making sure that there are enough of them left to satisfy hugepage reservations. Fix this by adding a return value to dissolve_free_huge_pages() and checking h->free_huge_pages vs. h->resv_huge_pages. Note that this may lead to the situation where dissolve_free_huge_page() returns an error and all free hugepages that were dissolved before that error are lost, while the memory block still cannot be set offline. Fixes: c8721bbb ("mm: memory-hotplug: enable memory hotplug to handle hugepage") Signed-off-by: Gerald Schaefer --- include/linux/hugetlb.h | 6 +++--- mm/hugetlb.c | 26 +++++++++++++++++++++----- mm/memory_hotplug.c | 4 +++- 3 files changed, 27 insertions(+), 9 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index c26d463..fe99e6f 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -450,8 +450,8 @@ static inline pgoff_t basepage_index(struct page *page) return __basepage_index(page); } -extern void dissolve_free_huge_pages(unsigned long start_pfn, - unsigned long end_pfn); +extern int dissolve_free_huge_pages(unsigned long start_pfn, + unsigned long end_pfn); static inline bool hugepage_migration_supported(struct hstate *h) { #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION @@ -518,7 +518,7 @@ static inline pgoff_t basepage_index(struct page *page) { return page->index; } -#define dissolve_free_huge_pages(s, e) do {} while (0) +#define dissolve_free_huge_pages(s, e) 0 #define hugepage_migration_supported(h) false static inline spinlock_t *huge_pte_lockptr(struct hstate *h, diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 603bdd0..91ae1f5 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1437,22 +1437,32 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed, /* * Dissolve a given free hugepage into free buddy pages. This function does - * nothing for in-use (including surplus) hugepages. + * nothing for in-use (including surplus) hugepages. Returns -EBUSY if the + * number of free hugepages would be reduced below the number of reserved + * hugepages. */ -static void dissolve_free_huge_page(struct page *page) +static int dissolve_free_huge_page(struct page *page) { + int rc = 0; + spin_lock(&hugetlb_lock); if (PageHuge(page) && !page_count(page)) { struct page *head = compound_head(page); struct hstate *h = page_hstate(head); int nid = page_to_nid(head); + if (h->free_huge_pages - h->resv_huge_pages == 0) { + rc = -EBUSY; + goto out; + } list_del(&head->lru); h->free_huge_pages--; h->free_huge_pages_node[nid]--; h->max_huge_pages--; update_and_free_page(h, head); } +out: spin_unlock(&hugetlb_lock); + return rc; } /* @@ -1460,16 +1470,22 @@ static void dissolve_free_huge_page(struct page *page) * make specified memory blocks removable from the system. * Note that this will dissolve a free gigantic hugepage completely, if any * part of it lies within the given range. + * Also note that if dissolve_free_huge_page() returns with an error, all + * free hugepages that were dissolved before that error are lost. */ -void dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn) +int dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn) { unsigned long pfn; + int rc = 0; if (!hugepages_supported()) - return; + return rc; for (pfn = start_pfn; pfn < end_pfn; pfn += 1 << minimum_order) - dissolve_free_huge_page(pfn_to_page(pfn)); + if (rc = dissolve_free_huge_page(pfn_to_page(pfn))) + break; + + return rc; } /* diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index b58906b..13998d9 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -1945,7 +1945,9 @@ static int __ref __offline_pages(unsigned long start_pfn, * dissolve free hugepages in the memory block before doing offlining * actually in order to make hugetlbfs's object counting consistent. */ - dissolve_free_huge_pages(start_pfn, end_pfn); + ret = dissolve_free_huge_pages(start_pfn, end_pfn); + if (ret) + goto failed_removal; /* check again */ offlined_pages = check_pages_isolated(start_pfn, end_pfn); if (offlined_pages < 0) { -- 2.8.4 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org